library(stringi)
library(stringr)

# Read play in using stri_read_lines()
earnest <- stri_read_lines("importance-of-being-earnest.txt")

# Detect start and end lines
start <- which(str_detect(earnest, pattern = "START OF THE PROJECT"))
end   <- which(str_detect(earnest, pattern = "END OF THE PROJECT"))

# Get rid of gutenberg intro text
earnest_sub  <- earnest[(start+1):(end-1)]

# Detect first act
lines_start <- which(str_detect(earnest_sub, pattern = "FIRST ACT"))
lines_start
# Set up index
intro_line_index <- 1:(lines_start-1)

# Split play into intro and play
intro_text <- earnest_sub[intro_line_index]
play_text <-  earnest_sub[-intro_line_index]

# Take a look at the first 20 lines
writeLines(head(play_text,n=20))

# Pattern for start word then .
pattern_1 <- START %R% one_or_more(WRD)%R% DOT
# Test pattern_1
str_view(play_lines, pattern = pattern_1, match = T)
str_view(play_lines, pattern = pattern_1, match = F)

# Pattern for start, capital, word then .
pattern_2 <- START %R% ascii_upper() %R% one_or_more(WRD) %R% DOT

# View matches of pattern_2
str_view(play_lines,pattern=pattern_2,match=T)

# View non-matches of pattern_2
str_view(play_lines,pattern=pattern_2,match=F)

# Get subset of lines that match
lines <- str_subset(play_lines,pattern=pattern_2)

# Extract match from lines
who <- str_extract(lines,pattern=pattern_2)

# Let's see what we have
unique(who)

str_view(lines,pattern=START %R%"Jack.")
str_subset(lines,pattern=START %R%"Jack.")
Jlines<-str_subset(lines,pattern=START%R%"Jack.")
length(Jlines)

JWords<-str_split(Jlines,pattern=fixed(" "))
WL<-sapply(JWords,length)
WL<-WL-1
mean(WL)

# Create vector of characters
characters <- c("Algernon", "Jack", "Lane", "Cecily", "Gwendolen", "Chasuble",
  "Merriman", "Lady Bracknell", "Miss Prism")

# Match start, then character name, then .
pattern_3 <- START %R% or1(characters) %R% DOT

# View matches of pattern_3
str_view(play_lines, pattern=pattern_3, match=T)

# View non-matches of pattern_3
str_view(play_lines, pattern=pattern_3, match=F)

# Pull out matches
lines <- str_subset(play_lines,pattern=pattern_3)

# Extract match from lines
who <- str_extract(lines,pattern=pattern_3)

# Unique characters
unique(who)

# Count lines per character
table(who)